This proposal in human population and quantitative genetics will provide powerful ways to investigate psychiatric disorders. Specifically to: (a) Develop models to analyze simultaneously in families a threshold character and a correlated quantitative measure (e.g., a diagnosis with a biochemical or psychosocial measure); (b) Implement multivariate analysis of quantitative traits in families with variable family structure, method of ascertainment, and incorporate indices into path analytic models to measure components of the liability for a disorder; (c) Extend multifactorial segregation analysis to half-siblings in order to detect genetic effects and to children in order to relate adult and child psychopathology; (d) Incorporate follow-up data into genetic models to study diagnostic unreliability and stability in a family study; (e) Incorporate variable age of onset into genetic models; and (f) Apply these techniques to several family data sets. Long term objectives are to develop genetic and non-genetic transmission models which are realistic representations of the familial distributions of psychiatric illnesses, develop techniques to test nosological hypotheses and resolve phenotypic heterogeneities, develop techniques to detect specific environmental and genetic factors, provide documented computer programs to the scientific community, and collaborate with others to study a broad spectrum of psychiatric and medical disorders. Methods include theoretical, computer implementation and simulation, and data analysis. First, theoretical issues in genetic epidemiology will be resolved, then path analytic and multivariate models will be extended and numerical techniques improved. Next, computer programs will be implemented and simulations used to assess operation characteristics. Finally, the methods will be applied to data. Genetic effects are important in many psychiatric disorders, but the ways in which they interact with the environment to produce an illness are unknown. This research will enable the definition of more etiologically homogeneous subgroups of affectd individuals so that underlying biochemical and psychosocial processes can be specified. In addition, specific modes of transmission can be characterized to improve risk prediction. This may lead to the discovery of relevant pre-morbid characteristics.